Technology · 1 sources covering this

OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue

The ChatGPT maker says its upcoming Astra model may have reached “critical” cyber capabilities, prompting it to halt a significant number of training runs while it tightens internal safeguards.

Share:X WhatsApp Reddit Facebook LinkedIn Open ↗

Coverage spectrum

1 L · 0 C · 0 R

How 1 sources are covering this

WIREDLean L·1h ago
OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue
The ChatGPT maker says its upcoming Astra model may have reached “critical” cyber capabilities, prompting it to halt a significant number of training runs while it tightens internal safeguards.
Read at WIRED →

Embed this bias spectrum on your site

Paste this iframe wherever you'd like the widget to appear. Resizes responsively; no JS required.

<iframe
  src="https://statesidedaily.com/embed/cluster/b7c2ad236cf8064b653e407e0fc55ebb4b119a91"
  width="100%" height="320"
  style="border:0;max-width:680px"
  loading="lazy"
  title="OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue"></iframe>

OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue

How 1 sources are covering this

OpenAI Overhauls Safety Protocols After Its AI Agents Went Rogue

More from Technology

Bitcoin Whales Add 43,000 Bitcoin in 60 Days

Expect Volatility In Results: Securitize CEO

FalconX CEO Talks Deals, Potential IPO, Layoffs

Spotify's Running Mode is now available on Android

Reader

Trust & legal

Across the network